⏱️ Skip Coverage After

By default, Early agents run coverage analysis twice during a test generation flow:

1. Before test generation — to establish a baseline
2. After test generation — to measure improvement and report results

The skip-coverage-after setting allows you to skip the second coverage run in order to reduce overall execution time.

---

🧠 What This Controls

Coverage analysis is used to:

• Decide which methods are eligible for test generation
• Measure coverage improvement
• Display before/after coverage metrics in the Early web application

When skip-coverage-after is enabled:

• Coverage is still calculated before test generation
• The post-generation coverage run is skipped
• The web application will display only the “before” coverage values

---

🔄 Default Behavior

By default:

• Coverage is executed both before and after test generation
• Coverage improvement is calculated and reported

This provides the most complete visibility, but adds additional runtime.

---

⚙️ Configuration Options

You can enable this optimization using either a CLI flag or an environment variable.

CLI Flag

--skip-coverage-after true | false

Environment Variable

export SKIP_COVERAGE_AFTER=true | false

Both options are equivalent.

---

🧪 Behavior Matrix

Setting	Coverage Before	Coverage After	Web Reporting
false (default)	✅ Yes	✅ Yes	Before & After coverage shown
true	✅ Yes	❌ No	Only Before coverage shown

---

🔧 When to Enable This

You may want to enable skip-coverage-after when:

• CI runtime needs to be minimized
• Running on large repositories or slow test suites
• Using Early in time-sensitive pipelines
• You care primarily about test generation, not immediate coverage delta reporting

This setting is particularly useful in combination with:

• MAX_TESTABLES
• CHANGED_METHODS_ONLY
• Reduced concurrency

---

🧪 Example: GitHub Actions

env:
  SKIP_COVERAGE_AFTER: "true"

This skips the post-generation coverage run for faster execution.

---

🧪 Example: Jenkins

environment {
  SKIP_COVERAGE_AFTER = 'true'
}

Or inline before running the CLI:

export SKIP_COVERAGE_AFTER=true
early generate-pr

---

📌 Notes

• Coverage is always required before test generation and cannot be skipped
• This setting affects reporting only — not test selection
• Skipping the second coverage run reduces accuracy of reported improvement
• The generated tests are not affected by this setting

---

✅ Summary

Setting	Default	Effect
--skip-coverage-after	false	Skip post-generation coverage analysis
SKIP_COVERAGE_AFTER	false	Environment-based equivalent
Affects	Reporting & runtime	No impact on test generation logic

---

Use this option when speed matters more than coverage delta visibility, especially in large or time-constrained CI workflows.